Observing Constraints - Cloud Cover Statistics

Mauna Kea

Data logged nightly by the UKIRT Telescope Operators over a ten-year period (1985 September 13 to 1996 August 4) have been analysed. Their reporting includes cloud cover (in eighths), cloud type and the number of usable and available hours. The data may be summarised thus:

Usable time: excluding nights for which no information was available (e.g. due to telescope shutdown) approximately three-quarters of the available time (31419 out of 43427 hr) was identified as "usable" i.e. the telescope was open and data were collected.

Cloudless: of the usable time 62% (19371 hr) was noted as being "cloudless". It is recognised that this is a subjective assessment e.g. it is difficult visually to detect thin cirrus in a dark sky. There are, however, two caveats: (a) we have included in this value only nights which were classified as cloudless throughout and (b) there may be a partial compensation from nights which were recorded as having some cloud cover (1/8 or 2/8, say), and which are treated as having these conditions all night, but which may have experienced substantial clear periods. Nevertheless, for the purposes of the Gemini observing condition constraint we have assumed that only 50% of the usable time is actually photometric.

Thin cloud: the UKIRT cloud cover and cloud type often were logged as a single value for an entire night. To estimate the time during which thin cloud was present, we have taken the nights explicitly reported as "thin cirrus" and added the fraction (1.0 - cloud cover) of nights reported as having "cirrus" (with a cloud cover of 4/8 or less). The UKIRT data shows these conditions occuring 23% (7096 hr) of the usable time. For the Gemini observing condition constraint we have assumed that thin cloud is present 20% of the usable time.

Thick cloud: for the Gemini observing condition constraint we have assumed that thick cloud is present 20% of the usable time.

Cerro Pachon

Although there is evidence that Cerro Pachon experiences significantly more photometric nights than Mauna Kea, at this time we have adopted the same values for the observing condition constraints.

Last update March 1, 2001; Phil Puxley